On 27-09-12 17:01, Raul Miller wrote:
Their variations seem to encode some domain knowledge, but what that knowledge
is is opaque to
me at the moment.
You are right. Proposition variants are dependent from others, e.g. if
you take hcr then you will miss the ones that have a pair in the first
house (fixed one: occupied by the Norwegian). I took some of the Prolog
embedded structures as a starting pointresulting in static proposition
variants opposed to the dynamic ones emerging in Prolog. Of course the
selector verb and the termination of the loop are crucial. Anyway I was
a bit surprised this method came with a solution (not hindered by too
much knowledge, but J shapes the mind is it?). Also you can not call
this a general solution, because you have to invest in enumerating legal
proposition variants.BTW other contributions on RC use all kind of hard
coded knowledge.Personally I prefer a language like Prolog to solve this
problem.
--
Met vriendelijke groet,
@@i = Arie Groeneveld
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
